Lecture 11: Perceptron

Content locked

The catalog is public. Reading this content requires approved access.

Introduction

Inspiration from Neuroscience

Linear Perceptron

Perceptron with Activation Function

Training the Perceptron

Perceptron Learning Algorithm

Comparison with Gradient Descent

Learning Logical Functions

AND Function

OR Function

NOT Function

XOR Problem (Minsky & Papert, 1969)

Activation Functions

Step Function

Sigmoid/Logistic Activation

Tanh Activation

ReLU Activation

Logistic Regression

Training: Maximum Likelihood Estimate

Training: Maximum A Posteriori (MAP)

Multilayer Perceptron

Single Hidden Layer Architecture

Detailed Example

Multilayer Perceptron Properties

Practice Problems

Brief History of Neural Networks

Key Takeaways